OXFAM INTERNATIONAL - GUMINE RESEARCH
Sorcery beliefs and practices in Gumine:
a source of conflict and insecurity
15 October 2010
Sorcery in Papua New Guinea 3
Profile of sorcery in Simbu Province and Gumine district 5
Study Sites 7
Ethical Consideration 8
RESEARCH FINDINGS 8
Beliefs about sorcery in Simbu 8
Identity of sorcerers 10
Impact of Sorcery accusations 11
Local solutions suggested by participants 14
The role of the police and judiciary 14
The role of the church 15
Oxfam International Papua New Guinea (PNG) through its peacebuilding and conflict reduction programme
in the Highlands of PNG conducts research that informs programme and policy developments. This paper is
an overview of research findings that were conducted as a part of the Security and Community Initiative
Research (SACIR) project in Gumine District of the Simbu Province, working with affected people to explore
their own situation, develop their own criteria of risk and determine local solutions to these problems. The
aim of the study was to explore how individuals and communities perceived insecurity, and how this
related to incidences of violence; looking at the nature of violence and conflict, common triggers and
impacts. At all sites that were visited, sorcery was often mentioned as a trigger and source of violence and
conflict in the community. This report will discuss those findings related to the practice of witchcraft and
sorcery and how this relates to insecurities faced in many families and the wider community in Gumine
District. The results are presented here in the context of recent literature on sorcery, in particular
important studies by the Health Services with IMR (Institute of Medical Research) and the Melanesian
Institute. It is hoped that this material will help to increase the understanding amongst Oxfam staff of the
nature and consequences of belief in sorcery and the role of the church and judiciary. The paper concludes
by suggesting ways in which Oxfam can work with its local partners to mitigate or reduce some of the
harmful consequences of sorcery beliefs.
Sorcery in Papua New Guinea
According to the Encyclopedia Britannica (1977), sorcery and witchcraft is defined as "ritual performance
that is thought to lead to the influencing of human or natural events by an external and impersonal
mystical force beyond the ordinary human sphere" (cited in Zocca and Urame 2008: 10). The majority of
Papua New Guineans do not accept natural causes as an explanation for misfortunes such as sickness,
accidents or death - instead these are attributed to supernatural causes which we have grouped under the
collective term of "Sorcery" 1 . Those accused of Sorcery are considered to have deliberately caused
misfortune through use of supernatural powers; they are usually punished by death, injury, destruction of
property or exile. Police reports show that victims have been buried alive, beheaded, choked to death,
thrown over cliffs or into rivers or caves, starved, axed, electrocuted, suffocated with smoke, forced to
drink petrol, stoned or shot (Amnesty International 2009).
In Papua New Guinea terms such as the pidgin sanguma, poison and witchcraft (together with a multitude of terms
in local languages) may be understood in different ways including: the possession of people by evil beings which take
animal form and which confer supernatural powers on their hosts (usually referred to as sanguma or kumo in Simbu);
forcing of a harmful object or substance into the victim so that they become sick or die, control of external powers
intrinsic to the person of the sorcerer which enable that person to inflict sickness or death by willpower alone; and
the ability to become invisible or to fly. Sorcery may involve contagious magic (malevolent actions on the 'leavings' of
a person), or involve causing direct physical harm to the victim. In this paper, as in Gibbs and Wailoni (2009), we have
decided to use the capitalized word Sorcery to cover all these, meaning the use of magical power to influence events.
There is a perception that sorcery related attacks are increasing in the Highlands, although this may be
partly due to a recent surge in media interest in the phenomenon 2 ; a lack of data on the past make true
comparisons difficult. The study by the Health Services and Institute of Medical Research (IMR) suggests
an increase in the number of sorcery-related attacks (Health Services and IMR 2004: 14). The Health
Services conducted an inquiry into the growing number of people suspected of being sorcerers presenting
to health services for treatment. They also note an increase in patients who believed their sickness was a
direct result of the practice of sorcery as well as relatives of those deceased seeking a post-mortem to
determine the cause of death. These phenomena were mostly recorded in EHP and Simbu province or
amongst those originating from Simbu.
Some argue that the apparent growing number of accusations of sorcery is a convenient disguise for
premeditated murder based more on a person's dislike for another, rivalry or revenge, rather than a deep-
rooted traditional belief of sorcery. There is also a concern that many murders may be linked to the
prevalence of HIV, the disease being seen as the result of the practice of witchcraft and sorcery magic
(Varolii 2010). The Health Services study attributes the increase in sorcery accusations in recent years to
high expectations that have turned into frustration and anxiety as many have found themselves excluded
from the benefits of development. The increase in sorcery cases may be a sign of this social stress (Health
Services and IMR 2004: 23). Historical studies looking at European witchcraft also link this phenomenon to
social change (Health Services and IMR 2004: 15).
The Health Services/IMR study looks at the changes that have taken place that may have contributed to a
change in sorcery accusations from an 'endemic' to an 'epidemic' state. Older people described changes in
killings of suspected sorcerers in recent years; they said that in the past, killings had the following
1. Occurred occasionally and only after careful consideration by elders that it was necessary for the
well-being of the community, now hearsay and rumor play a much larger role in accusations;
2. Never took place in public;
3. The body was never displaced for fear of the spirit entering another person;
4. No children or young people allowed to witness the killings;
5. Killing done with arrows or spears with weapons discarded with the body after the killing.
These characteristics no longer hold today, and the changes were said to have taken place after the 1980s.
It has also been suggested that in the past spirits and ancestors were more often blamed for misfortune,
but as these beliefs were suppressed by the church, accusations came to be focused instead on living
people (Lederman in Zocca 2009).
Another reason why sorcery must be of concern to organizations such as Oxfam is that the accused are
often the most vulnerable in society - for example in Simbu the majority of accused in the last few years
are women related to the deceased and aged between 40-60 years (Zocca 2009: 27) 3 . A report by Amnesty
International (2009:22) suggests that women are six times more likely to be accused of sorcery than men).
As we will see in this paper widows and the elderly are often targeted in Gumine.
The media reported the deaths of 50 women in 2009 through sorcery related attacks; according to the CEDAW 2010
PNG status report such cases have doubled in recent times (Varolii 2010).
The Minster of Community Development, Dame Carol Kidu in a statement to a UN committee on PNG's
progress to the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW) said
data collection on the number of attacks and deaths is lacking (Varolii 2010). A call for more research on
the issue was also made in a conference held in Madang in 2009 on 'Law on sorcery and sorcery-related
killings' (Gumar 2009). In Simbu, one of the earliest studies was in the 1950s when anthropologist Paula
Brown described sorcery in her book The Simbu: A Study of Change in the New Guinea Highlands (1972) as
a powerful weapon of revenge (Brown 1972 in Zocca 2009: 18). There have been several other books and
dissertations on sorcery in Simbu, particularly in the 1980s, but the most recent is the publication
'Sanguma in Paradise' (2009) by the Melanesian Institute in Goroka that looks at sorcery in both Simbu
Province many other parts of the country.
Profile of sorcery in Simbu Province and Gumine district
Simbu Province is made up of six districts; this study was conducted in the Gumine District situated in the
central part of the province. The reason for the choice of this district was the presence of Oxfam's partner,
Community Development Agency (see below). Mountainous and difficult terrain, high population density,
and a lack of economic opportunities have led to high levels of out-migration.
The part of the Melanesian Institute study which focuses on Simbu (Zocca 2009) looked at records kept in
hospitals, police stations and churches which then formed the basis for interviews to be carried out. Simbu
Province police records showed 67 Incidents relating to Sorcery from April 2000 to June 2005 in which 92
people were accused of Sorcery and were severely injured or killed (Zocca 2009: 21). According to these
records Gumine had the highest number of cases with 32 persons in total injured or killed in witchcraft or
sorcery related attacks (13 men and 19 women); of these, 26 died from their injuries. The majority of these
victims were in the 40-65 age range for both men and women. Kundiawa Hospital records supported this
with 75% of the cases presenting for treatment of sorcery-related injuries were elderly (Zocca 2009: 23).
Again, proportionally more females (31/48) than males presented (17/48). The police records suggest that
Gumine is the most affected district both in terms of the number of sorcery-related incidents and the
incidence of fatal attacks on the accused 4 . These official records show a slightly higher proportion of cases
affecting women rather than men, but interviews suggest that the disparity is in fact much greater, most
respondents suggesting that women bear the brunt of sorcery accusations and attacks.
Qualitative methodologies were used, including Participatory Rural Appraisal tools and techniques which
allowed violence and insecurity to be expressed in different ways by participants. The following diagramic
exercises were used:
Community Map - a graphical representation mapping the impact of insecurity on mobility. The mapping
exercise gave participants the opportunity to examine geographical areas of insecurity in their community
Gumine actually has the smallest population of any district in Simbu, so the high number of sorcery-related attacks
cannot be attributed to a high population.
and to identify areas for improvement. It was the first exercise to be done by participants, creating a
dialogue among participants and with the research team.
Venn Diagram - following the community mapping exercise the Venn diagram was used to explore types of
violence, the frequency with which they occurred and their impact on the lives of men and women
respectively. Circles representing each type of violence were drawn: the closer a circle to the community
circle in the centre, the more frequent the type of violence represented. The size of the circles indicated
the strength of the impact on the group concerned. The highlighted circles are the types of violence
affecting the group undertaking the exercise (men and women separately). An example is shown in Figure
Figure 1: Venn diagram of adult women from Dia (Kumai-Bomai LLG)
Problem Tree- is a problem analysis of the main triggers and corresponding effects of violence.
Peace Circle - was developed upon reflection and analysis of the problem tree for participants to form local
solutions that were practical and useful to the community.
Focus group discussions and in-depth interviews were
utilized for deeper analysis of the aspects of violence
and security that emerged through the diagramming
techniques. By using different techniques, we were
able to draw attention to diverse perspectives, and
confirm findings through triangulation. The
participatory nature of the study provided a voice to
groups vulnerable to violence and drew attention to a
broadened notion of security and to locally
appropriate intervention strategies. It allowed people
to develop their own criteria of risks and their own ideas about what appropriate interventions might look
like. Through the exercises participants were able to identify the root causes of problems and what needed
to be done to reduce the risk of violence. Participants stated that it helped to have people from outside
their community facilitate and guide them through the process so that they could critically reflect on their
The participants in PRA exercises formed four groups: adult men, adult women, young men and young
women. The exercises were conducted separately with each group, providing an opportunity for all to
voice their concerns and experiences. Each group consisted of 5-10 participants, however at times there
were more depending on the level of motivation and interest of the community to participate. Each group
identified a literate person who drew up the diagrams while the rest contributed ideas.
Participants made themselves available for the
research after notification by community
facilitators (see below). This was later followed
by a second briefing on the research by the
team upon arrival in the community.
Recruitment of participants for interviews was
done with the help of community facilitators,
leaders and spokes people in the community.
Often interviews would be carried out those
who had participated in the PRA exercises but
were still interested to go into further detail on
issues that concerned their community.
Oxfam supports the work of a local organisation, Community Development Agency (CDA) which has a
geographical focus on Gumine District. In those villages where the agency has a presence, it supports
community facilitators responsible for community mobilization and training. The research was conducted
in eight communities, including sites both with and without a CDA presence. The study covered locations
in all three Local Level Government (LLG) areas of the District. The communities are listed below:
Digine LLG - Kelmakeli and Gaima (two tribal enemies): Kelmakeli is situated up the mountain and is about
two hours walk from the main Gumine road. Kelmakeli has experienced election-related violence in 2007
with its neighboring tribe Gaima. Gaima is near the main Gumine highway. Gaima experienced tribal
violence with Kelmakeli during the 2007 national elections.
Kumai-Bomai LLG (three communities with no CDA presence): Dia community is about 30km from
Kundiawa town. It is located at a high altitude and is very cold almost throughout the day. Dia had no CDA
community facilitator at the time if the study, however the community was motivated and well organized.
The community also experiences tribal conflict during election periods and other issues, mostly related to
women, may cause tribal violence from time to time. Interviews were also conducted in two neighboring
communities: Diayuri and Gomgale.
Gumine LLG - (three non-tribal conflict communities): The Gumine highway passes through Kewaldien,
Bokoloma and Bokil villages. Tribal violence is rare and the communities are relatively peaceful, although
other types of violence were identified. In Kewaledian the last tribal fight was in 2001 which only lasted
two months. Bokolma is about ten minutes walk from Kewaledian. It is also near the Gumine highway and
experiences tribal conflict every 15-20 years. Bokil is about 15 minutes walk up the hill from Bokolma.
Tribal violence occurs once in a while, the last tribal fight was experienced in 2001. PRA exercises were
conducted in Bokil whilst interviews were conducted in Kawaladien and Bokolma.
Ethical consideration is very important when conducting research on insecurity and requires an extreme
sensitivity to the context and to the needs of the participants. Many participants have been traumatized by
many years of conflict and internal displacement. Others have suffered long term violence in marriage or
within the family. Consent was thus sought prior to interviews being carried out and confidentiality was
maintained and emphasized to participants throughout the interview.
Ethical considerations also demand that the findings should be presented back to the community.
Presentation of initial findings was duly made to each community, encouraging young people and women
to do the presentation. These events represented an opportunity for discussion and debate on main issues
that were identified through the research. Participants at all study sites confirmed that the data presented
was a true impression of their community.
Beliefs about sorcery in Simbu
Every society has its own explanations of the world and is interested about how and why things happen the
way they do. Simbu people like those in other parts of PNG and in the rest of Melanesia have a deep
commitment to family relations, strong connections with the land and beliefs in the spiritual world. These
beliefs and customs are passed through oral history: ancestral stories and myths promote this idea of a
non-empirical world that has an influence on the physical realm. The study showed that feelings of
insecurity were closely linked to this belief system. In the community mapping exercise, places of insecurity
were identified by participants. People fear both the spirits of the dead and locations associated with
them, including cemeteries and places where deaths had occurred such as battle-fields, rivers where
people had drowned or sites of intentional killings. Mountains, thick bush or forest and rivers or creeks
that have a particular significance or story attached to them are also places that people fear and avoid.
The main fear was of ancestral spirits often referred to as 'masalais', which may take human form and
which live in the mountains or bush away from human settlement. Belief in these spirits has implications
other than simple avoidance of certain sites: at one site that was visited, most men have two wives. If the
first wife tries to run away men may threaten to throw her clothes into the masalai stone, suggesting that
she will eventually die wherever she is. Men take advantage of this and have more than one wife. The
Health Services study mentions the belief that misfortunes caused by disturbing areas occupied by spirits
or masalais can either bring good fortune or bad (Health Services and IMR 2004:17-18).
Sorcery (usually referred to as sanguma) was one of the issues most commonly mentioned at all study
sites. Respondents expressed fear of the practice of sorcery and of perceived sorcerers who they believe
can directly harm themselves and their families, or bring about misfortune. The use of sorcery powers were
also linked to disasters occurring during warfare. For example, if a woman asks questions about how guns
are used, their provenance or purpose; and a man is subsequently killed in a fight, then his death will be
blamed on the woman. Women therefore, refrain from asking about the ownership and use of guns. They
keep well away from the affairs of men when it comes to tribal fights or guns, as it is seen to be men's
affair and questions will draw suspicion upon them.
According to Zocca (2009), in Simbu the most common form of destructive magic is known as kumo. This
term refers to an animal form which takes possession of a person and forces him/her to do evil things.
Despite this idea that the person practicing sorcery is possessed, they are still considered responsible for
the evil actions. Kumo is often used interchangeably with the pidgin word sanguma, which has a broader
meaning, and thus today kumo signifies both the intruding creatures and Sorcery in general (Zocca 2009:
20). 'Sangumas' are accused of causing illness, killing people, causing other misfortunes, feeding on the
flesh of dead people, feeding on human flesh, internal organs and waste, and taking the form of animals. It
is believed that the use of these powers is most likely to affect the immediate family. Those perceived to
possess those powers are despised and avoided.
The phenomenon of sorcery and witchcraft is an explanation used to understand unusual events, especially
those of a harmful nature (Health Services and IMR 2004: 15). 'Today hardly any death or sickness is
regarded as natural or accidental by the Simbus' (Zocca 2009:32). 'Abnormal' symptoms of illness or
sudden death, sickness and misfortune lead to suspicions of sorcery.
"We live close to each other and when one falls ill or a leader gets sick, they immediately accuse people
who they believe they practice sorcery. They burn their houses and attack them" (young woman,
People reflect on anything that the victim of the sorcery attack could have done that would have brought
on the misfortune. When there is an unexplained death in the village people start to ponder who might
have wanted the deceased dead, they remember their movements over the past few days and with whom
they may have come into contact. An example would be a person falling sick after taking food given by
someone else. In Gumine, young men said they only take food from their mothers because of this
insecurity. The accusation would be made against the person that cooked the food and gave it, usually a
female. A possible candidate is suggested and this then becomes hear-say; rumors gain momentum and,
with increased support from the community, they may lead to an attack. "Someone dies and then people
will start talking and then someone will start whispering and saying well you know, it might be this person"
(Father Phil Gibbs in McLeod 2004). Even when medical explanations are given relating to the cause of
death, there was always the question of, 'then what caused the illness or medical condition of the
deceased?' "A readiness to be deceived relieves people of the need to enguire further" (Health Services and
IMR 2004: 27).
When several people from the same family or community fall sick or die it is believed to be the work of
sorcerers. Such deaths may require more than one sorcerer and an agreement reached about the persons
to be killed. When people argue over land, they are attentive to the words used during the dispute. If later
one of the men dies, then those involved in the dispute may be accused of Sorcery. Any behaviour out of
the ordinary near the time of death can make a person a potential suspect of the death (Zocca 2009: 27).
The main motives behind supposed acts of sorcery include an outstanding debt not yet settled; feelings of
jealousy, envy or resentment of that person and anything that may have angered the sorcerer. Even when
an abnormally large number of livestock or domestic animals die it may be blamed on acts of sorcery. The
accused assert that accusations are made when people want to take over their land or possessions and
when there are existing disagreements between the parties.
"..for instance, if you have credit with other people and have not repaid it, there is a possibility that you
will be attacked. It is a big problem here and has resulted in fights between families, clan" (young
Innocent people can be accused of sorcery and killed out of jealousy or for revenge. Participants in Gumine
acknowledged that indeed, innocent people have been blamed and killed over the death of others. Those
that are accused of sorcery hold grudges against the people that have accused them, causing on-going
tensions and long term rifts within communities.
"Innocent people are being killed because of being suspected o/sanguma, while others have long lasting
conflicts" (young man, Kelawadian).
Identity of sorcerers
People harbor stereotypes about who practices sorcery and is thus in the possession of an evil
power. Although Sorcery accusations can be directed at both men and women particularly
elderly women are usually the first to be blamed and first to be targeted when there is an
unexplained death or misfortune that is experienced. In Gumine one women's group
stated that "when men die they put pressure on us the women and we are the first accused".
Respondents at all sites visited for this study said women were the first targets of sorcery
accusations when there was a death in the village. In Gumine adult women live with a constant fear of
being accused of sorcery. Some women mentioned that it is in fact men who practice sorcery, but manage
to avoid accusation by blaming women.
Other studies on Simbu found that sorcerers are often identified by their appearance and behavior. They look dirty
and have sores on their body, they are seen at odd hours at night, they behave strangely during funerals, look for
human waste, and they are always hungry for meat and stare at people when they are eating. The sorcerer would
have been seen around the place or area where the incident occurred. Weak and unimportant people are more likely
to be accused (Zocca 2009: 27).
"Sanguma related violence affects the mothers a lot because they are the main ones who are accused
when someone dies in the community" (adult women, Kelmakeli).
"The accusation of practicing witchcraft always leads to problems and troubles in the community.
Women especially the elderly are the common suspects of practicing sanguma" (adult woman,
Acts of sorcery are usually believed to occur within families: participants shared stories of being accused
and beaten by their own family members. One woman said that after her husband's death, she was
blamed for his death and suffered beatings at the hands of her own sons. Another woman stated that her
mother had been tortured and burnt by young men in their family after the death of her father. This
pattern is confirmed in the literature, which also suggests that accusations most often occur between close
relatives, creating disharmony, and family break-ups. The conflict and tension between the family and
parties involved is long-lasting and often people do not see face after accusations are made and even after
peace is brokered and some form of compensation paid for the shame.
Impact of Sorcery accusations
Torture, exile and murder
During the research respondents stated that sorcery related violence or killing occurs only occasionally and
not every time that a death occurs; it commonly follows deaths of young men because it is said that they
have a long life ahead of them and it has been cut short. Victims are tortured and interrogated sometimes
at gun-point to admit to the offence and to call the names of other sorcerers 6 . Accused persons said the
only way they could put an end to the torture was by admitting to the deed.
"When people are reluctant to say they are involved in Sanguma activities they are forced with gun to
confess that they are one of them" (young woman, Bokil).
This may involve testing whether the sorcerer does in fact the powers they profess to have by public
display. For example, telling the accused to empty the contents of a can of fish without it being opened. If
need be a witch-doctor or diviner called a 'glasman' is brought in to identify the sorcerer who caused the
death. Calling on glasman as a way to confirm suspicions demonstrates that 'proper procedures' have been
followed (Health Services and IMR 2004: 7). A witch-doctor may be brought in from another place: for
example, the relatives of one deceased person in Gumine sought an explanation by talking to a
witchdoctor in Madang, who then revealed the 'cause' of the death.
A number of respondents reported that, following confession or 'confirmation' of the suspect's guilt, they
are often beaten and banished from the community rather than killed and that, whilst sorcery-related
killings in the past were very high, these have been reduced now to casting out the accused 7 . Often the
Also recorded by Zocca (2009), who mentions that in Simbu, accusations of Sorcery may be confirmed (or refuted)
by responses during questioning: if the accused fails to answer appropriately according to the judgment of the
questioners, then they conclude that she is a Sorcery practitioner and they torture her or even kill her. Some
confessions are also exacted under torture.
It is difficult to judge whether this preference for banishment (noted in two communities) is real or increasing; Zocca
(2009:28) notes that in Simbu, there is preference to kill the accused rather than just banishing them from the
close relatives of the accused cannot bear the exile of their mothers or wives and leave also, resulting in
the effective departure of entire families. At one site visited about 20 households had left in this way. The
literature suggests that in Simbu sorcery-related displacement is a widespread phenomenon: one source
estimates that it has caused the dislocation of 10-15% of the Simbu population 8 . The displaced fall into four
main categories: those accused of sorcery practices who have been physically attacked and banished;
those who were accused and have run away for fear of their lives; supporters of the accused and lastly, a
group of (usually successful) people who have left their communities out of a fear of sorcery itself and the
destruction that those practicing it might bring to their property and wellbeing (Zocca 2009: 29-30).
Although banishment is certainly a common punishment meted out to those accused of sorcerery; injury
and death are also perpetrated on the accused.
"Sangumas are killed in different ways like thrown in the river, toilet, shot with gun and even burnt alive"
(young man, Kelawadian).
elated attacks recorded during this research included:
Burning with fire and hot iron rods
Cutting with knives
Beating with iron rods
Tying up and beating
Throwing into a river
Throwing off a cliff side (see Figure 3)
Banished from the community their house burnt and property destroyed
Buried alive (two elderly women and one elderly man)
Figure 2: A cliff from which people accused of practicing sorcery are thrown
"We kill people who are suspected of practicing
sorcery. They usually chop them up, drown them
in the river or even throw them off the cliff while
still alive" (adult man, Damagire).
"The people suspected of Sanguma have their
hands and legs tied up and put into a bag and
taken up to the mountain and rolled down the
hill. We have witnessed it" (young woman, Bokil).
community as sorcery can even be practiced from a distance, and the death itself may kill the kumo inside. Simbu.
Police records suggest that between 2000 and 2005 42% of all sorcery-related killings in the Province occurred in
Gumine district. One blogger reports that, according to Police records, half of all murder cases in 2008 in Simbu were
sorcery related, but this requires verification (Malum Nalu 2009).
8 Estimate by the Catholic bishop of Kundiawa (in Zocco 2009).
Wider violence in the community
In some instances the accused person's relatives retaliate with violence. These fights get bigger when the
accused has their supporters as well, dividing families and clans.
"If a house is closer to the accused sorcerer's house they will say you are the witchcraft and they will
burn his house and attack him. His people nearby will come in support of the accuser and retaliate and
the fight will get worse. This is what they do" (young woman, Bokolma).
These conflicts have a serious impact in terms of destruction of property. In 2006 one male respondent in
Gumine burnt down one of his brother's houses. At another site, the last tribal fight experienced was in
2001 and was a direct result of sorcery allegations: a woman was accused of killing her own husband by his
relatives, who attacked the woman's family, burning down houses and property. At yet another site, the
last incident of a sorcery-related killing (in 2002) also resulted in wider violence.
Sanguma leads to tribal fights, "especially when case is not mediated well or when people are not
satisfied with the mediation, they argue and fight. The fight gets worse, houses are burnt and people are
killed" (young woman, Bokolma).
Just calling another person a 'sanguma' can lead to violence. Those accused in this way are insulted,
causing anger and disunity among the family, clan or community.
"I was in Goroka with my husband to do some work when he called me and my mother witch (sanguma).
From this addressing, we had an argument which led to him compensating my mother" (adult women,
"If a woman calls another lady a sanguma, they fight over it. Support fight especially leading to big
fight" (young man, Gaima).
Factors which encourage sorcery-related violence: evidence from the literature
Many murders are committed in public so that other sorcerers are warned of the consequences of
practicing their magic, or using their powers. These acts are condoned by the community at large - they
encourage conformity and discourage unacceptable behavior - why then can't similar actions be carried
out on thieves or those who commit other wrongs? 9 The justification for barbarous acts committed against
sorcerers is that removal of the accused prevents further deaths and misfortune. One police officer in
Simbu stated "Once they eliminate the sorcerer they feel safe because they are saving other human beings
too. That's what they believe in" (in McLeod 2004). More than half the students in Simbu interviewed in
the Health Services study said the removal of a witch brought relief to the community. Less than half said it
increased fear or felt indifferent to the incident (Health Services and IMR 2004: 9). The perpetrators of the
torture and violence are usually young men. Young male perpetrators take the lead to attack the sorcerers
in an attempt to gain recognition and appreciation from the community (Zocca 2009: 29). The public stage
upon which acts of torture take place against sorcerers makes it difficult for those who might possibly try
Despite this, these acts do not necessarily deter crime (Patterson in Health Services and IMR, 2004: 15).
to intervene (Health Services and IMR 2004: 7-8). Few take the side of the accused or try to resolve the
problem, often through fear of association with the accused; such fears also limit the potential of better
educated people to influence the mindset of others.
Local solutions suggested by participants
During the study a peace circle exercise was used to that drew on local ideas to address current issues of
insecurity in villages. Here suggestions included the development of community laws prohibiting use of
violence against people accused of sorcery. At one site, young men interviewed who confessed to being
centrally involved in sorcery-related killings said they had only recently learnt about the sorcery act in the
media and would therefore refrain from inflicting violence on people accused of sorcery - they stated that
when someone dies there should not be any Sorcery accusations and insults. At other sites participants
responded by saying that acts of sorcery are spiritual in nature and must be brought to the attention of
pastors to deal with it in terms of praying and casting out that demon or evil spirit from those accused of
practicing sorcery. Others thought it should best be left to the law enforcement authorities to deal with.
Respondents at one site thought that the use of a witch-doctor or glasman should be made more
legitimate in order to put groundless suspicions to rest. At two sites many thought the banishment or
removal of those who were thought to practice sorcery out of the community was positive in that it
constituted a less violent approach to the matter than killing or torturing the accused.
None of the solutions suggested by participants called into question the existence of sorcery itself and
indeed only one participant in this study seemed to think that sorcery accusations might be groundless.
This was after he himself had been accused of sorcery. Now he tried to prevent other clan members from
making accusations and in doing so relates misfortune, death and sickness to natural causes.
The role of the police and judiciary
Local enforcement agencies, although aware of the enormity of the problem are not able to intervene. A
police officer in Kundiawa confirmed that it is difficult to intervene during an attack as a handful of
policemen cannot control a large mob of people who are determined to 'put right what is wrong' 10 . Most
cases involving sorcery murders are never reported and there is little cooperation in investigations carried
out to prosecute those implicated in the death. Other Government officials including Ward Councilors may
be better able to intervene. In May 2009 a fight would have erupted after the deaths of two children was
blamed upon two women and a man if it had not been for the local counselor who was able to respond
and stopped the conflict from escalating.
As mentioned above, there was some awareness of the Sorcery Act amongst respondents, but the stated
aim of the act is to 'to prevent and punish evil practices of sorcery and other similar evil practices'. Thus the
law explicitly acknowledges the 'existence' of sorcery and criminalizes those who practice it. Although, as
in any western-inspired court system proof is required, all parties concerned (including the court
magistrates) believe in sorcery, the 'proof provided by a diviner may be sufficient to send the accused to
prison (Longgar 2009). Use of the court system and sorcery act may however prevent summary executions
Comment made during visit of the UN rapporteur to PNG (at a meeting in Kundiawa attended by the author).
The Sorcery Act does have the merit of also criminalizing the torturers and killers of accused sorcerers. It is
stated in the preamble that there 7s a clanger that any law that deals fully with sorcery may encourage
some evil -intentioned people to make baseless or merely spiteful or malicious accusations that their
enemies are sorcerers solely to get them into trouble with other people, and this is a thing that the law
should prevent . But in reality penalties and provisions outlines in the Act focus mostly on the sorcerer as
perpetrator, and do not adequately cater for instances in which the alleged sorcerer is the victim which
presumably should be considered under legal charges of assault, manslaughter or murder. Zocca (2009:23)
reports that of the 67 Sorcery related incidents in Simbu recorded by the police between 2000 and 2005
(including 92 persons dead or injured) only six affidavits were produced regarding killers of suspected
witches and only two people were eventually sentenced.
On a national level the PNG Government had little to say when asked about what it was doing to stop the
murders of women accused of Sorcery when questioned by a UN inspection on women's rights in August
2010 in New York (Varolii 2010). In a visit by the UN Special Rapporteur to PNG (including Simbu) it was
stated that there was failure on the state to prevent and investigate sorcery accusations (Amnesty
International 2010: 7).
The role of the church
During this study, adult men in Gumine mentioned that church workers try to save people's lives and
prevent them from becoming victims of sorcery, by speaking out against people practicing sorcery and
driving fear into people (sorcerers) who practice such things. This suggests then that, for the church it is
not the accusers that are the problem, but those accused of sorcery. On the other hand it may have a
"Some religious people say it (sickness) is the Lords will but those who do not go to church captures
suspects and ask them guestions. If they believe they are Sanguma they throw them off the cliff"
(young woman, Bokil).
According to Zocca (2009), the approach of the catholic church in Simbu, lead by the Bishop Henk te
Maarssen has been countering the spread of witch hunts through an 8-point pastoral approach including
encouragement of the consideration of natural causes after a death, teaching about the causes of sickness,
meetings and courses on witchcraft, giving shelter to the accused, reporting to the police, and visits to
families of both accusers and accused. The diocese of Kundiawa has the policy of excluding from the
sacraments those accusing others of sorcery. However despite this position, the majority of Catholic
Church leaders believe that efforts must focus on stopping witches from continuing their evil practices.
Others stress attacking people's fear of witchcraft by convincing the faithful to rely on God's superior
power. In the case of other denominations the great majority of leaders now share fully in the beliefs of
their people - they therefore subscribe to the belief that sorcerers are in possession of an evil power which
enables them to kill someone, normally within blood relatives. The presentation of the Lutheran church at
the 'Sanguma in Paradise' seminar similarly reveals that for the district president of that church, the
'victims' are those who have been 'bewitched' (Zocca 2009).
A quotation from Zocco (2009) summarizes the impact of Sorcery in Simbu province:
"The fact that, in their majority, the accused and killed people in Simbu are elderly women or people of
low position in society is already a sign that something terribly wrong is going on in that society".
Perhaps the most terrible thing about such accusations is that, based on suspicion and hearsay, they
cannot be refuted by the accused. The usual instruments of criminal justice such as material evidence and
testimonials of witnesses cannot be brought to bear.
The community, so often a support mechanism, is the main perpetrator of this violence. Barbaric acts
sanctioned by the community are a reflection of the lack of awareness and understanding of two things:
• People do not understand or believe in the natural causes behind misfortune and thus attribute it
to sorcery, seeking to pin the blame on an individual.
• The 'evidence' used to implicate people in the practice of sorcery should be invalid in any legal
court, and yet is seen to be legitimate enough to justify extra-judicial killing, maiming and exile of
large numbers of people.
Accusations and fear of association with the accused are the crux of the problem. McLeod (2004) raises a
crucial issue with regards to any legal attempt to deal with sorcery attacks: 'prosecuting acts of violence is
possible, but how do you outlaw a genuine belief? 1
Education: The first and most important recommendation is to attack beliefs in Sorcery through the
education system, stressing instead an understanding of the real factors behind sickness and death.
Stressing the negative impacts of sorcery accusations on households, clans and communities in terms of
conflict, fear and suspicion can be done through plays and drama. There is a perception that sorcery beliefs
are part of the unique cultural heritage of Melanesia. Therefore it is important to stress that similar beliefs
exist, or have existed in the past, almost everywhere in the world, often with equally horrific
consequences. In Melanesia itself, beliefs have evolved over time and have perhaps not always been as
damaging as they are today.
Legal action: Although fear in sorcery overrides the fear of breaking the formal law, in the event that it is
believed that an act of sorcery has been made, parties concerned and the community should be
encouraged to take the matter to the appropriate courts to be settled under the Sorcery Act. Despite its
imperfections, this would at least avoid extra judicial killings and other forms of punishment without trial.
Reform of the sorcery act: The 1971 Sorcery Act presupposes the existence of sorcery by criminalizing
sorcerers as well as their torturers and killers; Zocca (2009) suggests that the law might be more effective if
it criminalized accusations of sorcery instead. There is also the question of diviners and those community
members whose support and approval make so many of the killings possible. Allegations and attacks
specifically aimed at female victims, highlights the need for the PNG Government to take action in terms of
its obligation under the CEDAW convention.
Next steps for Oxfam:
1. A partner workshop could be held covering the causes and consequences of sorcery beliefs, the
sorcery act and a comparison of different approaches to dealing with sorcery accusations. Staff at
the Melanesian Institute, some of whom are from Simbu themselves, have conducted workshops
on sorcery for church workers and could also be good facilitators of such events. The Institute is
putting together a book for facilitators which covers all the necessary topics and could be a useful
tool both directly for a partner workshop and for facilitators in communities.
2. In communities, partnership with the church could be an effective way of strengthening the impact
that Oxfam's partners can have on reduction of sorcery-related violence. The material in Zocca
(2009) suggests that the most progressive individuals are likely to belong to the Catholic Church
rather than the Pentecostal, Evangelical or Lutheran churches. The Bishop of Kundiawa in
particular may be a good point of contact when identifying the individuals concerned as he has
campaigned against sorcery accusation for decades.
3. Oxfam should facilitate KWP to build networks with sympathetic church leaders who are able to
identify and refer cases of sorcery to them: KWP could make victims of sorcery accusations one of
their primary targets as clients and put in place a support program which goes beyond counseling
to provide education, outreach and support for the displaced.
4. Oxfam needs to consider whether there is much that can be done at the national level in terms of
advocacy. Whilst reform of the sorcery act and better protection for those accused are essential, it
is not certain that Oxfam has channels through which these goals might be achieved. Recent
research has received widespread publicity, helped by a surge in media interest, although this
interest is not always to the benefit of the accused. The Melanesian Institute is currently working
together with the government's legal advisors on the sorcery act and has provided them with
recommendations for reform.
5. In terms of research, the large and recent studies by the Health Services and IMR and the
Melanesian Institute discussed in this report, suggest that further research into sorcery in general
is not needed; it will be important however for Oxfam to better understand the role and influence
of the church in the rural areas in which it works in order to initiate dialogue with these actors at
the community level. It would also be useful to know what is likely to happen to those accused of
sorcery who are actually taken to court: for example what types of 'evidence' are accepted by
courts in Gumine during sorcery trials and what proportion of those accused are actually charged
or condemned for under the sorcery act itself.
Amnesty International (2009). "PNG must act on sorcery killings" 10 July 2009, Amnesty International
Gibbs P. and Wailoni J. (2009). Sorcery and a Christian Response in the East Sepik. In: Sanguma in Paradise:
Sorcery, Witchcraft and Christianity in Papua New Guinea, Point No. 33, ed. Zocca, Franco. Melanesian
Institute for Pastoral and Socio-Economic Services Inc.
Gumar, P. (2009) Call for review of Sorcery Act, Malum Nalu 02 July 2009.
Health Services and Institute of Medical Research (2004). Final Report into the study of Sanguma in the
Eastern Highlands and Simbu Provinces. Health Services and Institute of Medical Research.
Longgar W. (2009). "Sorcery and Christianity in the Gazelle Peninsula" in Sanguma in Paradise. Sorcery,
Witchcraft and Christianity in Papua New Guinea, Point No. 33, ed. Zocca, Franco. Melanesian Institute for
Pastoral and Socio-Economic Services Inc.
McLeod, S. (2004). "Papua New Guinea - Sorcery" June 01 2004. Foreign Correspondent Australian
Broadcasting Commission, www.abc.net.au/foreign/contetn/2004/sll32881.htm
Paclii. (1998) Papua New Guinea Consolidated Legislation, Sorcery Act 1971. University of the South Pacific,
Per, Z. (2009) "Sorcery Act lacks bite: Law agencies", Malum Nalu 02 February 2009.
Rambo, K. and Brown, P. (1996). "Chimbu" Encyclopedia of World Cultures, 11 April 2010
Varolii, R. (2010). Papua New Guinea Apologizes for CEDAW Record, Women's eNews, 2 August 2010
Zocca, F. (2009). Witchcraft and Christianity in Simbu Province. In: Sanguma in Paradise. Sorcery,
Witchcraft and Christianity in Papua New Guinea, Point No. 33, ed. Zocca, Franco. Melanesian Institute for
Pastoral and Socio-Economic Services Inc.
Zocca, F and Urame, J. (2008). Sorcery, Witchcraft and Christianity in Melanesia. Melanesian Mission
Studies No. 5.